Every multi-location restaurant should track food cost percentage, labor cost percentage, prime cost, sales per labor hour, and inter-site variance, all reviewed weekly per location. The point is not collecting numbers, it is comparing sites against each other and against target so drift surfaces in days, not quarters. Prime cost under 60 percent is the single number that tells you if a location is healthy.
When you run one restaurant, you feel the numbers. When you run 14, you have to see them on a page, side by side, or you will never know that site 7 has been quietly bleeding labor for six weeks. Metrics for a multi-site operator are not a finance exercise. They are your eyes when you cannot be in the building.
The mistake almost every operator makes is the opposite of having no numbers. They have too many. A dashboard with 40 metrics is not visibility, it is noise, and noise gets ignored. The discipline of multi-unit operations is ruthless subtraction: find the handful of numbers that actually change what you do this week, put them on one page, and throw the rest out. Five numbers you act on beat 40 you admire.
There is a simple test for whether a metric earns a spot on your page. Ask: if this number moves, will I do something different this week? If the answer is no, the metric is trivia, no matter how interesting it looks on a chart. Average check size is interesting, but it rarely changes your Monday. Prime cost on a struggling site changes your Monday completely, because it tells you exactly which manager gets a call. Keep the numbers that drive an action and cut the ones that only drive a conversation.
The Five Numbers That Actually Drive Decisions
You can drown in dashboards. Most owners track 40 metrics and act on none. The discipline is picking the few that change what you do this week. These five do, and each one earns its place by pointing at a specific lever you can pull.
- Food cost percentage, the margin on what you sell and the first place waste and over-portioning show up
- Labor cost percentage, the biggest controllable expense and the one most sensitive to scheduling discipline
- Prime cost, food plus labor, the true health gauge that is hard to fake by starving one side
- Sales per labor hour, productivity per person on the floor and the check on whether cheap labor is actually understaffing
- Inter-site variance, the gap between your best and worst location and the clearest map of recoverable margin
Targets You Can Hold Sites Against
A metric without a target is trivia. Each of these needs a number every location is measured against, so a struggling site is obvious the moment the report prints.
| Metric | Why it matters | Target |
|---|---|---|
| Food cost % | Margin on product sold | 28-35% |
| Labor cost % | Largest controllable cost | 25-35% |
| Prime cost % | Overall operating health | Under 60% |
| Sales per labor hour | Floor productivity | Over $60 |
| Inter-site variance | Consistency across locations | Under 4 pts |
Prime Cost Is the One Number to Rule Them All
If you only watch one figure, watch prime cost: food plus labor as a percentage of sales. It captures your two biggest controllables in a single line, and it is hard to fake. A location can hide a food-cost problem by cutting labor, but prime cost catches both. Hold every site under 60 percent and you have a fighting chance at real profit.
Across my 14 locations, ranking by prime cost every week told me instantly who needed a call. The bottom two sites got 15 minutes of attention. The rest got left alone to run. That is the whole value of metrics: they tell you where to point your limited time.
That last point is the one operators underrate. Metrics are not really about the numbers. They are about your attention, which is the scarcest resource you have across a multi-unit group. You cannot walk every line every day, so the numbers walk it for you and report back where the trouble is. A site running a clean 56 percent prime cost does not need you this week, and trusting the number to tell you that is what frees you to spend your time where it actually changes an outcome. Good metrics do not just measure the business. They ration your attention to the spots that pay you back for spending it.
How the Five Numbers Catch a Problem the Average Hides
Here is why you watch five numbers and not one blended figure. Imagine site 9 posts a healthy 58 percent prime cost. Looks fine. But underneath, food cost is running 38 percent and labor is running an unusually low 20 percent. The prime cost looks good only because the manager is cutting labor to mask a food-cost disaster, and an understaffed kitchen is about to wreck ticket times and quality. One number lied. The component numbers told the truth. That is the case for keeping food and labor visible alongside prime cost, not just the rolled-up total.
- Read prime cost first to see overall health at a glance
- Split it into food and labor to see which controllable is driving the number
- Check sales per labor hour to confirm the floor is actually productive, not just cheap
- Compare the site against its own trend and against the group
- Flag any number that looks good only because another number is being starved
Sales per labor hour is the metric most operators leave off the page, and it is the one that catches the labor version of the same trick. A site can post a beautiful labor percentage by simply cutting bodies until the number looks good, but a kitchen that is too lean is a kitchen that misses ticket times, drops quality, and burns out the people who stayed. Sales per labor hour tells you whether the labor on the floor is actually productive or just thin. A low labor cost with low sales per labor hour is not efficiency, it is understaffing dressed up as a win, and it shows up a few weeks later as bad reviews and turnover. Read the two numbers together and the masquerade collapses.
Review Weekly, Compare Sites, Coach the Laggards
Numbers reviewed monthly are history. Reviewed weekly, they are a steering wheel. The cadence is what makes the metrics work, not the metrics themselves.
- Pull the five metrics per location every week onto one page
- Rank sites worst to best on prime cost
- Call the bottom two and find the single root cause
- Assign one fix with an owner and a date
- Re-check next week and celebrate the climbers
Want help turning these numbers into a weekly system your managers actually run? Start with a free audit.
Book a free auditThe Mistake of Measuring Without Comparing
Plenty of operators track these five numbers and still get no value from them, because they look at each site in isolation. A 33 percent food cost looks acceptable until you see the identical concept across town running 29. The power of a multi-unit operation is the built-in benchmark: every site is a controlled experiment on the same menu and the same systems, so the spread between them is not just information, it is a target. When your best site proves a number is reachable inside your own walls, your worst site has no excuse left. Do not just measure. Rank, compare, and chase the spread to zero.
The Bottom Line
Multi-location metrics exist to give you visibility you lose the moment you cannot walk every line yourself. Track five numbers, hold every site to a target, and review weekly so the gaps between locations cannot hide. Start with prime cost on one page this week. The operators who scale are the ones who turned their gut feel into a number everyone can see.